Referral Limit Policy for the Credit Authorization Process

This paper considers the credit authorization problem in credit card companiespsila authorization systems. Customers are classified into several risk segments according to the type of services requested and the associated credit risk of customers. Card companies need to balance the risk exposure and the quality of customer service in the decision of referral/no referral. We formulate this problem as a Markov decision process (MDP) and characterize the optimal policy as referral limit control for each risk segment. A numerical example is provided to verify our results.